136 Bni.I. SySTIlM TECIIMC.il lOlliX.II. 



raising the IM). bfiwccii tlic ])l;ilcs sulVuiciitly, the locus of zero 

 field can be driven hark into coinridenre with the emitting plate; 

 beyond whieh stage, the "limitation of current by space-charge" 

 ceases. But if the P.I), is sufficiently low the potential minimum (or 

 maximum) is prominent and is remote from the electrode, and in 

 these cases the equations we have just deduced are inapplicable. 



It thus may readily happen that when we apply a certain potential 

 to one electrode and a certain other potential to another electrode 

 separated from the first one by gas or vacuum, we may find points 

 between them where the potential is not ittlerniediate between the 

 potentials of the electrodes. This is a (jueer conclusion, to an\hody 

 accustomed to the How of electricity in wires, liul ii is true, and 

 must be kept in mind. 



0. Tiiii Self-M.m.ntai.ninu Disc h.vrcus 



The Arc ought to be the easiest to understand among the self- 

 maintaining discharges, in one rcsjicct at least; for it keeps its own 

 cathode so inten.sel>- hot that thermionic electrons are supplied con- 

 tinuously in great abundance at the negative end of the discharge, 

 and the theorist can begin his labors l)y trying to explain how and 

 why this high temperature is maintained. Anything which lends to 

 lower the temperature of the cathode, for instance b>- draining heat 

 awa>- from it, is \er>- perilous to the arc. Stark uses various schemes 

 for preventing the cathotle from growing \ery hot, and they all killed 

 the arc. This alsf) ex()lains why the arc is most difficidl to kindle 

 and most inclined to flicker out when formed between electrodes of 

 a metal which conducts heat exceptionally well, and most durable 

 when formed between electrodes of carbon, which is a comparati\ely 

 poor conductor for heal. It probably explains why the arc has a 

 harder time to keep itself alive in hydrogen, a gas of high thermal 

 conductivity, than in air. While the gas in which the arc has its 

 being and the anode to which it extends both influence the discharge, 

 the high temperature of the cathode is cardinal. 



The cathode is presumably kept hf)t by the rain of jKhsitive ions 

 upon it, striking it with violence and yielding up their energy of 

 motion to it; at least this is the obvious and plausible explanation. 

 Now the arc is commonly and easily maintained in fairly dense gases, 

 with a com|)aratively small potential-difference between widely- 

 si'parated electrodes; and the euerg\- which an ion can acquire from 

 the field strength prevailing in it, in the short interval between two 

 collisions with nrnki uUs, ij. so sm.ill lli.it it cannot be made to account 



